Structural Crack Monthly Width Measurement & Propagation Log

ContextThe substrate governs what an acceptable crack width looks like. In brick masonry, a 0.3 mm crack at a mortar joint is common shrinkage behaviour; in reinforced concrete, 0.3 mm may signal rebar corrosion requiring immediate investigation because chlorides and moisture have reached the steel. For render and plaster, fine cracks under 0.2 mm are usually cosmetic. Timber-framed walls crack differently to load-bearing masonry. Misidentifying the substrate leads to incorrect severity grading and either unnecessary alarm or dangerous complacency. Note the substrate once and update only if the finish layer has been replaced since last month.

ContextOrientation reveals the load causing the crack. Vertical cracks in brick walls often indicate differential settlement or unrestrained thermal expansion. Horizontal cracks in retaining walls or basement walls are serious — they suggest lateral soil pressure is overcoming structural resistance and warrant immediate engineer review regardless of width. Diagonal or stepped cracks in masonry typically indicate differential foundation movement. A crack that shifts from vertical to diagonal between monthly inspections signals changing stress patterns — this is more significant than a crack that is widening but maintaining a consistent angle. Record the angle on the location map, not just in the written log.

ContextIsolated cracks behave differently from mapped cracking (a network of fine random cracks across a surface) or parallel cracking (multiple cracks running in the same direction). Mapped cracking in concrete often indicates alkali-silica reaction (ASR), a progressive chemical deterioration with no simple repair. Parallel horizontal cracks in brickwork at regular courses may indicate corrosion of bed-joint reinforcement, which eventually causes spalling. Crazing on render is usually cosmetic shrinkage. Documenting the pattern type each month lets you detect when an isolated crack begins spawning satellite cracks — a key escalation trigger that would be invisible if you only measured width.

ContextPlace your crack comparator card directly against the crack face in good lighting, matching the crack to the closest line on the card. The widest point is your primary datum and typically occurs mid-length or at a stress concentration such as a corner or window reveal. If using a digital crack gauge, press the caliper feet firmly against the crack edges without rocking. Record to the nearest 0.05 mm. Eyeballing width without a comparator introduces errors of up to 0.5 mm — enough to completely miss a significant month-on-month change. Bright oblique lighting from a torch held at a shallow angle improves crack visibility on pale or smooth surfaces.

ContextA crack does not open uniformly along its length. By measuring at three defined points — both ends and the centre — you build a profile showing whether the crack is widest at its root (suggesting the driving force originates at that end) or at its tip (suggesting recent propagation). Label these points P1 (top or left end), P2 (midpoint), P3 (bottom or right end) and mark them on the close-up photograph. Consistent point selection each month is essential — measuring even 20 mm from your previous point can introduce a 0.1 mm apparent change in a tapered crack where no real movement has occurred.

ContextCompare today's widest-point reading against last month's. A change of less than 0.05 mm falls within measurement uncertainty and should be logged as "no detectable change." An opening of 0.05–0.1 mm per month is classified as slow and should continue to be monitored. More than 0.1 mm opening in a single month warrants a written flag in the log and verbal notification to the responsible engineer. Any crack showing consistent opening across three or more consecutive months must be escalated regardless of its absolute width — the trend matters more than the current number. A closing crack is not automatically safe; a crack that opened 0.4 mm over four months and then closed 0.05 mm is still 0.35 mm worse than baseline.

ContextUse a flexible steel rule or follow the crack path with a measuring tape pressed against the surface. Do not measure a straight line between end-markers if the crack meanders, as this underestimates length by up to 30% on sinuous cracks and misrepresents the area of material under stress. Record in millimetres. Total length is a secondary propagation indicator: a crack that is not widening but is extending rapidly is still active and should be flagged. Length data also feeds into structural engineer assessments of stress distribution along a wall or slab.

ContextEnd-markers — a dated paint stroke or a tell-tale tip positioned exactly at each crack terminus — are your most reliable propagation indicators. If the crack has grown beyond last month's marker, measure the extension length and record the direction of growth (for example: "extended 12 mm southeast toward window reveal"). Direction matters because cracks propagating toward structural elements such as lintels, columns, or bearing points change risk category faster than cracks propagating into open wall field. Add a new dated marker at the current terminus after measuring, ensuring the old marker remains visible for continuity.

ContextA crack branching into two or more paths indicates increasing stress or a change in the material's failure mode. Branching in reinforced concrete can signal that the crack is deflecting around aggregate or that rebar at the branching point has debonded. Satellite cracks appearing near the main crack tip often precede a rapid length increase in the following month. Record the branch angle and its distance from the nearest measurement point on the main crack. Every satellite crack must be assigned its own code and added to the location map immediately — treating it as part of the parent crack creates ambiguity that complicates trend analysis.

ContextTell-tales (crack monitors) measure movement in two axes: across the crack (opening or closing) and along the crack face (shear or sliding movement). Read both axes and record to 0.5 mm for manual tell-tales or 0.1 mm for digital extensometers. The X-axis reading tracks crack opening and closing; the Y-axis tracks vertical or horizontal shear. Shear movement is often more structurally significant than pure opening because it indicates differential vertical displacement across the crack — a sign that the two sides of the crack are not simply breathing thermally but moving in different directions under load. If your installed tell-tale only provides one axis, note this limitation explicitly in the log.

ContextA tell-tale that has been bumped, painted over, or partially debonded from the substrate is unusable and must be marked as invalid. Look for: offset alignment of the red and white grid that does not match the cumulative readings (suggesting a knock since last month), paint bridging across the monitor face (which physically restrains movement and makes all subsequent readings false), and adhesive failure visible as a gap behind the monitor's edges. If a tell-tale has been disturbed, mark the reading as "invalid — reset required," replace or reinstall it, and note the interruption in the data chain. Never backfill an invalid reading with an estimate or interpolation — a documented gap is more defensible than a fabricated value.

ContextDemec points are small stainless steel discs bonded to concrete, used with a precision Demec gauge to measure compressive or tensile strain between two fixed points. Check that each disc is firmly bonded (push gently — any rocking indicates debonding), that the seating hole is free of paint or filler, and that the disc has not corroded to the point where the gauge pin will not seat fully. A missing Demec point cannot be replaced mid-series without resetting the zero datum and losing historical comparability. Record any loss immediately and flag to the engineer of record so they can advise on re-establishing the baseline measurement series.

ContextInsert the thinnest feeler blade that enters the crack without force, then step up in thickness until resistance is felt. Record the depth as "> X mm" (meaning the crack is at least that deep) or as "through" if the blade passes completely. Surface cracks under 3 mm deep in render are usually cosmetic. Cracks deeper than 10 mm in masonry or concrete have penetrated the structural material and alter the load path through the wall. Through-wall cracks in below-grade or retaining walls are especially serious because they allow water ingress that accelerates deterioration through freeze-thaw cycling and rebar corrosion. Never widen a crack to probe it — doing so damages the evidence and may affect structural performance.

ContextA crack visible on one face may be a surface-only fracture or may have propagated through the full wall thickness. Inspect the opposite face for any crack at the corresponding plan position, even a hairline. Documenting it as "single-face," "both faces — offset" (crack has deviated through the thickness), or "both faces — aligned" (straight through-crack) changes the severity classification. Aligned through-cracks in load-bearing masonry double the urgency because the wall has lost continuity across its full section. In cavity walls, both inner and outer leaves must be inspected independently as the cavity can hide the relationship between the two faces.

ContextCrack widths change with temperature — concrete expands approximately 0.01 mm per metre of wall length per degree Celsius — and with moisture content, as masonry and timber swell in high humidity. A reading taken in January at 2°C will differ measurably from one taken in July at 28°C even in a perfectly stable structure. Without this data logged in the same row as the measurements, you cannot distinguish seasonal movement from progressive structural deterioration when you review trend data six months later. The thermometer and hygrometer must be read at the inspection location, not taken from an online weather station, because indoor and sheltered surfaces can differ significantly from ambient conditions.

ContextEvents that can accelerate or explain cracking include: prolonged drought causing shrinkage of clay soils and foundation settlement, heavy sustained rainfall following a dry period causing rapid ground heave, ground frost causing ice lens formation in porous masonry, nearby excavation or driven piling causing vibration and soil de-stressing, and significant new loads added to the structure such as plant equipment, added floors, or altered water tanks. These events explain anomalous readings and prevent false escalation — or false reassurance if movement appears to have stopped temporarily. If no notable events occurred, log "no significant events" rather than leaving the field blank. A blank is ambiguous and creates doubt during later review.

ContextMany buildings show a predictable annual movement cycle: cracks widen in cold months as materials contract and narrow in summer as they expand. Plotting monthly readings against average monthly temperature over a 12-month period reveals whether a crack's movement is purely seasonal or carries a progressive component. A crack that returns to exactly the same width each July for three years is almost certainly seasonal movement requiring monitoring but not intervention. One that is 0.1 mm wider each July despite the same seasonal cycle is progressing — even if the month-to-month delta appears small. This analysis requires at least 13 consecutive months of data to be valid and should be performed as a formal annual summary.

ContextThe scale reference is non-negotiable for photographic evidence. Use a bright yellow or red scale card showing at least 100 mm of graduation, positioned parallel to the crack and in contact with the wall surface — not floating in mid-air — so that perspective distortion does not misrepresent scale. The card and crack must both be in focus; if your phone struggles with shallow depth of field, use a wide aperture or tap-to-focus on the crack rather than the scale. Disable HDR, portrait mode, and any edge-softening filters — crack sharpness is diagnostic, and software smoothing can eliminate hairline cracks or obscure width detail. Blurry or scale-free photographs are inadmissible in formal structural assessments and insurance claims.

ContextAdopt a consistent convention such as "NW04_P2_2026-06-08.jpg" — crack code, measurement point, ISO date. Storing photos in date-named folders without a crack code makes retrieval extremely slow once you have accumulated 200 or more images over a monitoring period. This naming convention also enables automated comparison tools and AI image processing software to sort and analyse your archive without manual intervention. Cloud-backup each session before leaving the site. A single dropped or stolen phone can destroy an entire monitoring record that took years to build. If cloud backup is not available on-site, copy files to a USB drive kept separately from the phone.

ContextThe map should show every crack code, its orientation as a drawn line (not just a dot), the latest total length, the location and code of each tell-tale, and any new cracks discovered this month. Use a consistent scale and orientation — north always up for floor plans, elevation name always in the header for wall elevations. If a crack has branched or a satellite crack has appeared, update the map immediately during the inspection rather than from memory at the office. A well-maintained, current location map is often the most persuasive document when briefing a structural engineer, presenting to an insurer, or handing over the monitoring programme to a new inspector.

ContextMonthly change equals this month's widest-point reading minus last month's widest-point reading. Record it with an explicit sign: "+0.15 mm" (opening), "−0.05 mm" (closing), "0.00 mm" (no measurable change). Running totals — the cumulative change from the first measurement date — are more revealing than month-to-month deltas alone and should appear as a separate column. A crack that has opened a net 1.2 mm over 12 months of small increments is significantly worse than a crack that opened 0.3 mm suddenly and then stopped, yet the month-to-month deltas might look similar. Both numbers are needed to tell the full story.

ContextA 0.1 mm per month rate in a 3 mm crack represents a 3.3% monthly increase — measurable, meaningful, and too fast to ignore without review. Flag these entries prominently in both the paper log and any digital record. The flag creates the obligation to review, not to assume structural failure: seasonal cracks near external walls may legitimately exceed this threshold in cold months. What the flag does is force a written judgement call — either confirm the movement is within expected seasonal range and document the reasoning, or escalate. An unflagged threshold exceedance that later leads to damage is the worst outcome: it suggests the monitoring programme was not actually being read.

ContextA single anomalous reading can reflect instrument error, measurement-point drift, or a one-off weather event. Three consecutive months of movement in the same direction strongly indicates a real and ongoing process. In your trend analysis section of the log, mark any such crack as "progressively active." This classification should trigger a written notification to the building owner or structural engineer of record, regardless of the absolute width. The classification must be reviewed if movement subsequently reverses.

ContextUse a consistent classification scale every month without deviation so that category changes are meaningful. A widely adopted framework for masonry, adapted from BRE Digest 251, runs as follows: Category 0 is hairline under 0.1 mm — aesthetic only, no action; Category 1 is fine cracks 0.1 to 1 mm — redecoration needed; Category 2 is slight cracks 1 to 5 mm — point-filling needed; Category 3 is moderate cracks 5 to 15 mm or several greater than 3 mm — engineer to assess; Category 4 is severe cracks 15 to 25 mm — engineer to inspect urgently; Category 5 is very severe cracks greater than 25 mm — potential structural failure, immediate action required. Record the category this month and note any category change from last month — a step from Category 1 to Category 2 must be highlighted even if the absolute width is still small.

ContextThe following conditions require escalation regardless of your internal severity framework: any crack wider than 5 mm in a load-bearing wall, column, or beam; any through-wall crack in a retaining wall, basement wall, or below-grade structure; any crack showing greater than 0.3 mm of movement in a single month; any crack accompanied by door or window binding that was not present last month; any crack that is wet, showing active seepage, or covered in white efflorescence (salt deposits indicating ongoing moisture movement through the crack); and any new crack appearing in a location with no previous cracking history. Escalation does not mean the building is unsafe — it means the assessment has exceeded the scope of monitoring and requires engineering judgement to interpret.

ContextThe log must be signed by the person who physically performed the inspection — not the report author if they are different individuals. Include the inspector's full name, qualifications if applicable, the inspection date and start time, and a one-line weather conditions summary. A signed log is a legal document in many regulatory and insurance contexts. Store the original in a fireproof location or in a cloud system with version control and access logging. Provide a copy to the building owner or facilities manager on the same day, not at the end of the month. Logs should be retained for a minimum of 10 years, or for the life of the structure if it is a critical or heritage building — whichever is longer.